Cytochrome P450s are a large superfamily of enzymes that primarily hydroxylate substrates using dioxygen, although other redox-type reactions, including some reductions, have been reported. One variant, cytochrome P450 BM-3 is found in the bacterium Bacillus megaterium (EC 1.14.14.1). This variant, also known as CYP102, is a water-soluble, catalytically self-sufficient P450 containing a monooxygenase domain (64 kD) and a reductase domain (54 kD) in a single polypeptide chain (Narhi and Fulco, Journal of Biological Chemistry, 261 (16): 7160-7169 (1986) and Journal of Biological Chemistry, 262 (14): 6683-6690 (1987); Miura and Fulco, Biochimica et Biophysica ACTA, 388 (3): 305-317 (1975); Ruettinger et al., 1989). The minimum requirements for activity of the BM-3 variant are substrate, dioxygen and the cofactor nicotinamide adenine dinucleotide phosphate (NADPH). Nucleotide and amino acid sequences for P450 BM-3 can be found in, and are hereby incorporated by reference from, the GenBank database under the accession Nos. J04832 (SEQ ID NO: 1) and P14779 (SEQ ID NO: 2), respectively.
P450 BM-3 hydroxylates fatty acids of chain lengths between C12 and C18 at subterminal positions, and the regioselectivity of oxygen insertion depends on the chain length (Miura and Fulco, Biochimica et Biophysica ACTA 388 (3): 305-317 (1975); Boddupalli et al., Journal of Biological Chemistry 265 (8): 4233-4239 (1990)). The natural substrates of P450 BM-3 are hydroxylated at their ω-1, ω-2, and ω-3 positions using atmospheric dioxygen and nicotinamide adenine dinucleotide phosphate (NADPH) as shown in FIG. 1. (Ost et al., Biochemistry, 40, 13430-13438 (2001)). Substrate is bound and hydroxylated in a hydrophobic binding pocket that is positioned directly above a heme cofactor which is located in its own domain of the protein. A single peptide chain connects this heme domain to the reductase domain of the protein where NADPH is reduced and flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD) cofactors are used to transfer electrons to the heme active site for catalysis. The resulting products of the catalysis can be seen in FIG. 1. The hydroxylation of myristic acid by cytochrome P450 BM-3 results in 53.6% ω-1 hydroxylation product, 24.5% ω-2 hydroxylation product, and 20.0% ω-3 hydroxylation product. However, none of these substrates of P450 BM-3 are alkanes.
The optimal chain length of saturated fatty acid substrates for P450 BM-3 is 14-16 carbons, and the enzyme was initially believed to have no activity towards fatty acids smaller than C12 (Miura and Fulco, Biochimica et Biophysica ACTA, 388 (3): 305-317 (1975)). The activity of P450 BM-3 on saturated fatty acids follows the order C15=C16>C14>C17>C13>C18>C12 (Oliver et al., Biochemical Journal, 327: 537-544 Part 2 (1997)). On the C16 fatty acid, kcat=81 s−1 and Km=1.4×10−6 M (kcat/Km=6.0×107 M−1s−1). With the C12 fatty acid, kcat=26 s−1, Km=136×0−6 M and kcat/Km=1.9×05 M−1s−1 (Oliver et al., Biochemical Journal, 327: 537-544 Part 2 (1997)). P450 BM-3 is also known to hydroxylate the corresponding fatty acid amides and alcohols and forms epoxides from unsaturated fatty acids (Miura and Fulco, Biochimica et Biophysica ACTA, 388 (3): 305-317 (1975); Capdevila et al., J. Biol. Chem. 271:22663-22671 (1996); Graham-Lorence et al., J. Biol. Chem., 272:1127-1135 (1997); Ruettinger and Fulco, Journal of Biological Chemistry, 256 (11): 5728-5734 (1981)). The enzyme was reported to be inactive towards alkanes and methyl esters lacking the polar functionality of the natural substrates (Miura and Fulco, Biochimica et Biophysica ACTA, 388 (3): 305-317 (1975)). However, there were indications that P450 BM-3 could accept shorter-chain alkanes, although with very low activity (Munro et al., Biochem Soc Trans, 21 (4): 411S (1993)). However, wild type BM-3 was ineffective in its ability to hydroxylate alkanes, as the turnover of the enzyme was less than 100 total, and the rate was reported to be at 80 min−1.
Additionally, relative to other enzymes that hydroxylate linear alkanes, wild type BM-3 was also ineffective. For example, Pseudomonas oleovorans is able to oxidize n-alkanes using hydroxylase machinery comprising an integral membrane oxygenase (omega-hydroxylase), a soluble NADH-dependent reductase and a soluble metalloprotein (rubredoxin) which transfers electrons from the reductase to the hydroxylase (Staijen et al., European Journal of Biochemistry, 267 (7): 1957-1965 (2000)). The omega-hydroxylase has been cloned from P. oleovorans into E. coli, where it has been expressed and purified (Shanklin et al., Proceedings of the National Academy of Sciences of the United States of America, 94 (7): 2981-2986 (1997)). The specific activity of this omega-hydroxylase for octane (5.2 units/mg hydroxylase (Shanklin et al., Proceedings of the National Academy of Sciences of the United States of America, 94 (7): 2981-2986 (1997)) is about 13 times greater than that of P450 BM-3 (0.4 units/mg enzyme). (The specific activity of the complete P. oleovorans system, including the rubredoxin and the reductase, is less than 5.2 units/mg). Thus, wildtype P450 BM-3 was inefficient relative to this (and other) naturally occurring enzymes for alkane hydroxylation.
While the wild-type P450 was found to be ineffective in alkane hydroxylation, this inefficiency has been overcome in previous work by one of the Inventors. In this work, directed evolution was used to convert wild type BM-3 into a fast, but non-selective, alkane hydroxylase, dubbed “139-3.” (Farinas et al., Adv. Synth. Catal., 343, 601-606 (2001); Glieder et al., Nature Biotech., 20, 1135-1139 (2002)). The P450 139-3 was found to have an increased oxidation activity towards alkanes, and was found to be active on alkanes as small as propane. In comparison to the P450 BM-3, the evolved 139-3 protein has 11 amino acid substitutions in its heme domain.